An effective AIDS vaccine must protect against sexual transmission of HIV. We have developed a model of sexually transmitted aids using vaginal infection of female rhesus macaques with SIV for studies of pathogenesis and evaluation of vaccines. We previously showed that direct inoculation of the vaginal submucosa (perivaginal inoculation) with a live-attenuated SIV, SIVMAC1a11, induces immune responses that are similar in quality to those of virulent virus. Such immune responses may reduce virus load, but may not be of sufficient strength or duration to prevent infection by virulent virus inoculated intravaginally. To determine whether SIV-specific CTL precursors could be re-stimulated after replication of SIVMAC1a11 wanes, we compared two perivaginal immunization regimens. SIVMAC1a11 was the first immunization; the second immunization was given 8 weeks later. Group a (n = 2) received a second perivaginal inoculation with SIVMAC1a11. To try to selectively stimulate TH1-type CD4+ helper T cells that may augment long term memory for SIV-specific cytotoxic T lymphocyte (CTL) responses, group B (n = 6) animals received a second perivaginal inoculation with whole inactivated SIV without any adjuvant. All animals were transiently viremic and had detectable secondary CTL activity to SIV env and gag antigens after immunization with live SIVMAC1a11. Nine weeks after the booster, all immunized animals and 4 naive controls were challenged intravaginally with cell-free SIVMAC251. Two weeks after challenge two immunized (group b) and two naive animals were necropsied to assay virus load and tissue distribution as well as cell-mediated immune responses in genital and peripheral lymphoid tissues. Because women are often exposed to relatively low doses of HIV by sexual contact, a vaccine that could generate genital immune responses that significantly reduced levels of virus transmitted during vaginal exposure might prevent or delay systemic infection.